Evaluating Potential Distribution of High‐Risk Aquatic Invasive Species in the Water Garden and Aquarium Trade at a Global Scale Based on Current Established Populations

Aquatic non‐native invasive species are commonly traded in the worldwide water garden and aquarium markets, and some of these species pose major threats to the economy, the environment, and human health. Understanding the potential suitable habitat for these species at a global scale and at regional scales can inform risk assessments and predict future potential establishment. Typically, global habitat suitability models are fit for freshwater species with only climate variables, which provides little information about suitable terrestrial conditions for aquatic species. Remotely sensed data including topography and land cover data have the potential to improve our understanding of suitable habitat for aquatic species. In this study, we fit species distribution models using five different model algorithms for three non‐native aquatic invasive species with bioclimatic, topographic, and remotely sensed covariates to evaluate potential suitable habitat beyond simple climate matches. The species examined included a frog (Xenopus laevis), toad (Bombina orientalis), and snail (Pomacea spp.). Using a unique modeling approach for each species including background point selection based on known established populations resulted in robust ensemble habitat suitability models. All models for all species had test area under the receiver operating characteristic curve values greater than 0.70 and percent correctly classified values greater than 0.65. Importantly, we employed multivariate environmental similarity surface maps to evaluate potential extrapolation beyond observed conditions when applying models globally. These global models provide necessary forecasts of where these aquatic invasive species have the potential for establishment outside their native range, a key component in risk analyses.     

Risk Analysis for Invasive Species: General Framework and Research Needs

A joint workshop was convened by the Society for Risk Analysis Ecological Risk Assessment Specialty Group and the Ecological Society of America Theoretical Ecology Section to provide independent scientific input into the formulation of methods and processes for risk assessment of invasive species. In breakout sessions on (1) the effects of invasive species on human health, (2) effects on plants and animals, (3) risk analysis issues and research needs related to entry and establishment of invasive species, and (4) risk analysis issues and research needs related to the spread and impacts of invasive species, workshop participants discussed an overall approach to risk assessment for invasive species. Workshop participants agreed on the need for empirical research on areas in which data are lacking, including potential invasive species, native species and habitats that may be impacted by invasive species, important biological processes and phenomena such as dispersal, and pathways of entry and spread for invasive species. Participants agreed that theoretical ecology can inform the process of risk assessment for invasive species by providing guidelines and conceptual models, and can contribute to improved decision making by providing a firm biological basis for risk assessments.     

Risk Assessment for Invasive Species

Although estimates vary, there is a broad agreement that invasive species impose major costs on the U.S. economy, as well as posing risks to nonmarket environmental goods and services and to public health. The domestic effort to manage risks associated with invasive species is coordinated by the National Invasive Species Council (NISC), which is charged with developing a science-based process to evaluate risks associated with the introduction and spread of invasive species. Various international agreements have also elevated invasive species issues onto the international policy agenda. The World Trade Organization (WTO) Sanitary and Phytosanitary (SPS) Agreement establishes rights and obligations to adhere to the discipline of scientific risk assessment to ensure that SPS measures are applied only to the extent required to protect human, animal, and plant health, and do not constitute arbitrary or unjustifiable technical barriers to trade. Currently, however, the field of risk assessment for invasive species is in its infancy. Therefore, there is a pressing need to formulate scientifically sound methods and approaches in this emerging field, while acknowledging that the demand for situation-specific empirical evidence is likely to persistently outstrip supply. To begin addressing this need, the Society for Risk Analysis Ecological Risk Assessment Specialty Group and the Ecological Society of America Theoretical Ecology Section convened a joint workshop to provide independent scientific input into the formulation of methods and processes for risk assessment of invasive species to ensure that the analytic processes used domestically and internationally will be firmly rooted in sound scientific principles.     

Risk Analysis for Biological Hazards: What We Need to Know about Invasive Species

Risk analysis for biological invasions is similar to other types of natural and human hazards. For example, risk analysis for chemical spills requires the evaluation of basic information on where a spill occurs; exposure level and toxicity of the chemical agent; knowledge of the physical processes involved in its rate and direction of spread; and potential impacts to the environment, economy, and human health relative to containment costs. Unlike typical chemical spills, biological invasions can have long lag times from introduction and establishment to successful invasion, they reproduce, and they can spread rapidly by physical and biological processes. We use a risk analysis framework to suggest a general strategy for risk analysis for invasive species and invaded habitats. It requires: (1) problem formation (scoping the problem, defining assessment endpoints); (2) analysis (information on species traits, matching species traits to suitable habitats, estimating exposure, surveys of current distribution and abundance); (3) risk characterization (understanding of data completeness, estimates of the "potential" distribution and abundance; estimates of the potential rate of spread; and probable risks, impacts, and costs); and (4) risk management (containment potential, costs, and opportunity costs; legal mandates and social considerations and information science and technology needs).     

Optimal Strategies for Interception,Detection, and Eradication in Plant Biosecurity

The introduction of invasive species causes damages from the economic and ecological point of view. Interception of plant pests and eradication of the established populations are two management options to prevent or limit the risk posed by an invasive species. Management options generate costs related to the interception at the point of entry, and the detection and eradication of established field populations. Risk managers have to decide how to allocate resources between interception, field detection, containment, and eradication minimizing the expected total costs. In this work is considered an optimization problem aiming at determining the optimal allocation of resources to minimize the expected total costs of the introduction of Bemisia tabaci‐transmitted viruses in Europe. The optimization problem takes into account a probabilistic model for the estimation of the percentage of viruliferous insect populations arriving through the trade of commodities, and a population dynamics model describing the process of the vector populations' establishment and spread. The time of field detection of viruliferous insect populations is considered as a random variable. The solution of the optimization problem allows to determine the optimal allocation of the search effort between interception and detection/eradication. The behavior of the search effort as a function of efficacy or search in interception and in detection is then analyzed. The importance of the vector population growth rate and the probability of virus establishment are also considered in the analysis of the optimization problem.     

Knowledge-Based Risk Assessment Under Uncertainty for Species Invasion

Management of invasive species depends on developing prevention and control strategies through comprehensive risk assessment frameworks that need a thorough analysis of exposure to invasive species. However, accurate exposure analysis of invasive species can be a daunting task because of the inherent uncertainty in invasion processes. Risk assessment of invasive species under uncertainty requires potential integration of expert judgment with empirical information, which often can be incomplete, imprecise, and fragmentary. The representation of knowledge in classical risk models depends on the formulation of a precise probabilistic value or well-defined joint distribution of unknown parameters. However, expert knowledge and judgments are often represented in value-laden terms or preference-ordered criteria. We offer a novel approach to risk assessment by using a dominance-based rough set approach to account for preference order in the domains of attributes in the set of risk classes. The model is illustrated with an example showing how a knowledge-centric risk model can be integrated with the dominance-based principle of rough set to derive minimal covering "if ... , then...," decision rules to reason over a set of possible invasion scenarios. The inconsistency and ambiguity in the data set is modeled using the rough set concept of boundary region adjoining lower and upper approximation of risk classes. Finally, we present an extension of rough set to evidence a theoretic interpretation of risk measures of invasive species in a spatial context. In this approach, the multispecies interactions in an invasion risk are approximated with imprecise probability measures through a combination of spatial neighborhood information of risk estimation in terms of belief and plausibility.     

Characterizing the Performance of the Conway‐Maxwell Poisson Generalized Linear Model

Count data are pervasive in many areas of risk analysis; deaths, adverse health outcomes, infrastructure system failures, and traffic accidents are all recorded as count events, for example. Risk analysts often wish to estimate the probability distribution for the number of discrete events as part of doing a risk assessment. Traditional count data regression models of the type often used in risk assessment for this problem suffer from limitations due to the assumed variance structure. A more flexible model based on the Conway‐Maxwell Poisson (COM‐Poisson) distribution was recently proposed, a model that has the potential to overcome the limitations of the traditional model. However, the statistical performance of this new model has not yet been fully characterized. This article assesses the performance of a maximum likelihood estimation method for fitting the COM‐Poisson generalized linear model (GLM). The objectives of this article are to (1) characterize the parameter estimation accuracy of the MLE implementation of the COM‐Poisson GLM, and (2) estimate the prediction accuracy of the COM‐Poisson GLM using simulated data sets. The results of the study indicate that the COM‐Poisson GLM is flexible enough to model under‐, equi‐, and overdispersed data sets with different sample mean values. The results also show that the COM‐Poisson GLM yields accurate parameter estimates. The COM‐Poisson GLM provides a promising and flexible approach for performing count data regression.     

Risk analysis of fast spreading species in a Kashmir Himalayan National Park (Dachigam) for better monitoring and management

Huge economic costs and ecological impacts of invasive alien species (IAS) in the protected areas (PAs) worldwide make their timely prediction and potential risk assessment of central importance for effective management. While the preborder weed risk assessment framework has been extensively evaluated and implemented, the postborder species risk assessment framework has not been subjected to the same degree of scrutiny. Here we used a rather more realistic modified version of the Australian Weed Risk framework (AWRM) for Dachigam National Park (DNP) in Kashmir Himalaya against 84 plant species, including 55 alien species and 29 fast spreading native species, for risk analysis. We found two very high-risk species, three high-risk species, 10 medium-risk species, 29 low-risk species, and 40 negligible-risk species in the DNP. The containment scores accordingly ranged from 14.4 to 293.5 comprising of 27 species that can be contained with very high feasibility, 23 species with high feasibility, 14 species with medium feasibility, and 12 species which cannot be contained easily thereby having low feasibility of containment (FOC) score. However, eight species which have a negligible FOC score are difficult to contain within their infestation sites. Our results demonstrate the merit of the AWRM with a caution that the necessary region-specific modifications may help in its better implementation. Overall, these results provide quite a promising tool in the hands of protected area managers to timely and effectively deal with the problem of plant invasions.     

Regression Versus Causation,Revisited

I recently discussed pitfalls in attempted causal inference based on reduced‐form regression models. I used as motivation a real‐world example from a paper by Dr. Sneeringer, which interpreted a reduced‐form regression analysis as implying the startling causal conclusion that “doubling of [livestock] production leads to a 7.4% increase in infant mortality.” This conclusion is based on: (A) fitting a reduced‐form regression model to aggregate (e.g., county‐level) data; and (B) (mis)interpreting a regression coefficient in this model as a causal coefficient, without performing any formal statistical tests for potential causation (such as conditional independence, Granger‐Sims, or path analysis tests). Dr. Sneeringer now adds comments that confirm and augment these deficiencies, while advocating methodological errors that, I believe, risk analysts should avoid if they want to reach logically sound, empirically valid, conclusions about cause and effect. She explains that, in addition to (A) and (B) above, she also performed other steps such as (C) manually selecting specific models and variables and (D) assuming (again, without testing) that hand‐picked surrogate variables are valid (e.g., that log‐transformed income is an adequate surrogate for poverty). In her view, these added steps imply that “critiques of A and B are not applicable” to her analysis and that therefore “a causal argument can be made” for “such a strong, robust correlation” as she believes her regression coefficient indicates. However, multiple wrongs do not create a right. Steps (C) and (D) exacerbate the problem of unjustified causal interpretation of regression coefficients, without rendering irrelevant the fact that (A) and (B) do not provide evidence of causality. This reply focuses on whether any statistical techniques can produce the silk purse of a valid causal inference from the sow's ear of a reduced‐form regression analysis of ecological data. We conclude that Dr. Sneeringer's analysis provides no valid indication that air pollution from livestock operations causes any increase in infant mortality rates. More generally, reduced‐form regression modeling of aggregate population data—no matter how it is augmented by fitting multiple models and hand‐selecting variables and transformations—is not adequate for valid causal inference about health effects caused by specific, but unmeasured, exposures.     

Categorical Regression Analysis of Acute Exposure to Tetrachloroethylene1

Exposure-response analysis of acute noncancer risks should consider both concentration (C) and duration (T) of exposure, as well as severity of response. Stratified categorical regression is a form of meta-analysis that addresses these needs by combining studies and analyzing response data expressed as ordinal severity categories. A generalized linear model for ordinal data was used to estimate the probability of response associated with exposure and severity category. Stratification of the regression model addresses systematic differences among studies by allowing one or more model parameters to vary across strata denned, for example, by species and sex. The ability to treat partial information addresses the difficulties in assigning consistent severity scores. Studies containing information on acute effects of tetrachloroethylene in rats, mice, and humans were analyzed. The mouse data were highly uncertain due to lack of data on effects of low concentrations and were excluded from the analysis. A model with species-specific concentration intercept terms for rat and human central nervous system data improved fit to the data compared with the base model (combined species). More complex models with strata denned by sex and species did not improve the fit. The stratified regression model allows human effect levels to be identified more confidently by basing the intercept on human data and the slope parameters on the combined data (on a C × T plot). This analysis provides an exposure–response function for acute exposures to tetrachloroethylene using categorical regression analysis.     

Nonparametric Tree‐Based Predictive Modeling of Storm Outages on an Electric Distribution Network

This article compares two nonparametric tree‐based models, quantile regression forests (QRF) and Bayesian additive regression trees (BART), for predicting storm outages on an electric distribution network in Connecticut, USA. We evaluated point estimates and prediction intervals of outage predictions for both models using high‐resolution weather, infrastructure, and land use data for 89 storm events (including hurricanes, blizzards, and thunderstorms). We found that spatially BART predicted more accurate point estimates than QRF. However, QRF produced better prediction intervals for high spatial resolutions (2‐km grid cells and towns), while BART predictions aggregated to coarser resolutions (divisions and service territory) more effectively. We also found that the predictive accuracy was dependent on the season (e.g., tree‐leaf condition, storm characteristics), and that the predictions were most accurate for winter storms. Given the merits of each individual model, we suggest that BART and QRF be implemented together to show the complete picture of a storm's potential impact on the electric distribution network, which would allow for a utility to make better decisions about allocating prestorm resources.     

Predicting the Risk of Biological Invasions Using Environmental Similarity and Transport Network Connectedness

Understanding the risk of biological invasions associated with particular transport pathways and source regions is critical for implementing effective biosecurity management. This may require both a model for physical connectedness between regions, and a measure of environmental similarity, so as to quantify the potential for a species to be transported from a given region and to survive at a destination region. We present an analysis of integrated biosecurity risk into Australia, based on flights and shipping data from each global geopolitical region, and an adaptation of the “range bagging” method to determine environmental matching between regions. Here, we describe global patterns of environmental matching and highlight those regions with many physical connections. We classify patterns of global invasion risk (high to low) into Australian states and territories. We validate our analysis by comparison with global presence data for 844 phytophagous insect pest species, and produce a list of high‐risk species not previously known to be present in Australia. We determined that, of the insect pest species used for validation, the species most likely to be present in Australia were those also present in geopolitical regions with high transport connectivity to Australia, and those regions that were geographically close, and had similar environments.     

First Application of FISK,the Freshwater Fish Invasiveness Screening Kit,in Northern Europe: Example of Southern Finland

The climatic conditions of north temperate countries pose unique influences on the rates of invasion and the potential adverse impacts of non‐native species. Methods are needed to evaluate these risks, beginning with the pre‐screening of non‐native species for potential invasives. Recent improvements to the Fish Invasiveness Scoring Kit (FISK) have provided a means (i.e., FISK v2) of identifying potentially invasive non‐native freshwater fishes in virtually all climate zones. In this study, FISK is applied for the first time in a north temperate country, southern Finland, and calibrated to determine the appropriate threshold score for fish species that are likely to pose a high risk of being invasive in this risk assessment area. The threshold between “medium” and “high” risk was determined to be 22.5, which is slightly higher than the original threshold for the United Kingdom (i.e., 19) and that determined for a FISK application in southern Japan (19.8). This underlines the need to calibrate such decision‐support tools for the different areas where they are employed. The results are evaluated in the context of current management strategies in Finland regarding non‐native fishes.     

A Meta‐Analysis of Carbon Nanotube Pulmonary Toxicity Studies—How Physical Dimensions and Impurities Affect the Toxicity of Carbon Nanotubes

This article presents a regression‐tree‐based meta‐analysis of rodent pulmonary toxicity studies of uncoated, nonfunctionalized carbon nanotube (CNT) exposure. The resulting analysis provides quantitative estimates of the contribution of CNT attributes (impurities, physical dimensions, and aggregation) to pulmonary toxicity indicators in bronchoalveolar lavage fluid: neutrophil and macrophage count, and lactate dehydrogenase and total protein concentrations. The method employs classification and regression tree (CART) models, techniques that are relatively insensitive to data defects that impair other types of regression analysis: high dimensionality, nonlinearity, correlated variables, and significant quantities of missing values. Three types of analysis are presented: the RT, the random forest (RF), and a random‐forest‐based dose‐response model. The RT shows the best single model supported by all the data and typically contains a small number of variables. The RF shows how much variance reduction is associated with every variable in the data set. The dose‐response model is used to isolate the effects of CNT attributes from the CNT dose, showing the shift in the dose‐response caused by the attribute across the measured range of CNT doses. It was found that the CNT attributes that contribute the most to pulmonary toxicity were metallic impurities (cobalt significantly increased observed toxicity, while other impurities had mixed effects), CNT length (negatively correlated with most toxicity indicators), CNT diameter (significantly positively associated with toxicity), and aggregate size (negatively correlated with cell damage indicators and positively correlated with immune response indicators). Increasing CNT N₂‐BET‐specific surface area decreased toxicity indicators.     

A Diversity Index for Model Space Selection in the Estimation of Benchmark and Infectious Doses via Model Averaging

In chemical and microbial risk assessments, risk assessors fit dose‐response models to high‐dose data and extrapolate downward to risk levels in the range of 1–10%. Although multiple dose‐response models may be able to fit the data adequately in the experimental range, the estimated effective dose (ED) corresponding to an extremely small risk can be substantially different from model to model. In this respect, model averaging (MA) provides more robustness than a single dose‐response model in the point and interval estimation of an ED. In MA, accounting for both data uncertainty and model uncertainty is crucial, but addressing model uncertainty is not achieved simply by increasing the number of models in a model space. A plausible set of models for MA can be characterized by goodness of fit and diversity surrounding the truth. We propose a diversity index (DI) to balance between these two characteristics in model space selection. It addresses a collective property of a model space rather than individual performance of each model. Tuning parameters in the DI control the size of the model space for MA.     

Spatial Stochastic Simulation Offers Potential as a Quantitative Method for Pest Risk Analysis

Pest risk analysis represents an emerging field of risk analysis that evaluates the potential risks of the introduction and establishment of plant pests into a new geographic location and then assesses the management options to reduce those potential risks. Development of new and adapted methodology is required to answer questions concerning pest risk analysis of exotic plant pests. This research describes a new method for predicting the potential establishment and spread of a plant pest into new areas using a case study, Ralstonia solanacearum, a bacterial disease of potato. This method combines current quantitative methodologies, stochastic simulation, and geographic information systems with knowledge of pest biology and environmental data to derive new information about pest establishment potential in a geographical region where a pest had not been introduced. This proposed method extends an existing methodology for matching pest characteristics with environmental conditions by modeling and simulating dissemination behavior of a pest organism. Issues related to integrating spatial variables into risk analysis models are further discussed in this article.     

A New Multicriteria Risk Mapping Approach Based on a Multiattribute Frontier Concept

Invasive species risk maps provide broad guidance on where to allocate resources for pest monitoring and regulation, but they often present individual risk components (such as climatic suitability, host abundance, or introduction potential) as independent entities. These independent risk components are integrated using various multicriteria analysis techniques that typically require prior knowledge of the risk components’ importance. Such information is often nonexistent for many invasive pests. This study proposes a new approach for building integrated risk maps using the principle of a multiattribute efficient frontier and analyzing the partial order of elements of a risk map as distributed in multidimensional criteria space. The integrated risks are estimated as subsequent multiattribute frontiers in dimensions of individual risk criteria. We demonstrate the approach with the example of Agrilus biguttatus Fabricius, a high‐risk pest that may threaten North American oak forests in the near future. Drawing on U.S. and Canadian data, we compare the performance of the multiattribute ranking against a multicriteria linear weighted averaging technique in the presence of uncertainties, using the concept of robustness from info‐gap decision theory. The results show major geographic hotspots where the consideration of tradeoffs between multiple risk components changes integrated risk rankings. Both methods delineate similar geographical regions of high and low risks. Overall, aggregation based on a delineation of multiattribute efficient frontiers can be a useful tool to prioritize risks for anticipated invasive pests, which usually have an extremely poor prior knowledge base.     

Tree‐Based Logistic Regression Approach for Work Zone Casualty Risk Assessment

This study presents a tree‐based logistic regression approach to assessing work zone casualty risk, which is defined as the probability of a vehicle occupant being killed or injured in a work zone crash. First, a decision tree approach is employed to determine the tree structure and interacting factors. Based on the Michigan M‐94\I‐94\I‐94BL\I‐94BR highway work zone crash data, an optimal tree comprising four leaf nodes is first determined and the interacting factors are found to be airbag, occupant identity (i.e., driver, passenger), and gender. The data are then split into four groups according to the tree structure. Finally, the logistic regression analysis is separately conducted for each group. The results show that the proposed approach outperforms the pure decision tree model because the former has the capability of examining the marginal effects of risk factors. Compared with the pure logistic regression method, the proposed approach avoids the variable interaction effects so that it significantly improves the prediction accuracy.     

A Decision Support Tool for Predicting Patients at Risk of Readmission: A Comparison of Classification Trees,Logistic Regression,Generalized Additive Models,and Multivariate Adaptive Regression Splines

The number of emergency (or unplanned) readmissions in the United Kingdom National Health Service (NHS) has been rising for many years. This trend, which is possibly related to poor patient care, places financial pressures on hospitals and on national healthcare budgets. As a result, clinicians and key decision makers (e.g., managers and commissioners) are interested in predicting patients at high risk of readmission. Logistic regression is the most popular method of predicting patient‐specific probabilities. However, these studies have produced conflicting results with poor prediction accuracies. We compared the predictive accuracy of logistic regression with that of regression trees for predicting emergency readmissions within 45 days after been discharged from hospital. We also examined the predictive ability of two other types of data‐driven models: generalized additive models (GAMs) and multivariate adaptive regression splines (MARS). We used data on 963 patients readmitted to hospitals with chronic obstructive pulmonary disease and asthma. We used repeated split‐sample validation: the data were divided into derivation and validation samples. Predictive models were estimated using the derivation sample and the predictive accuracy of the resultant model was assessed using a number of performance measures, such as area under the receiver operating characteristic (ROC) curve in the validation sample. This process was repeated 1,000 times—the initial data set was divided into derivation and validation samples 1,000 times, and the predictive accuracy of each method was assessed each time. The mean ROC curve area for the regression tree models in the 1,000 derivation samples was .928, while the mean ROC curve area of a logistic regression model was .924. Our study shows that logistic regression model and regression trees had performance comparable to that of more flexible, data‐driven models such as GAMs and MARS. Given that the models have produced excellent predictive accuracies, this could be a valuable decision support tool for clinicians (healthcare managers, policy makers, etc.) for informed decision making in the management of diseases, which ultimately contributes to improved measures for hospital performance management.